Update app.py

app.py

@@ -1,9 +1,10 @@
 import subprocess
-subprocess.run(["pip", "install","gradio","torch","transformers"])
+subprocess.run(["pip", "install", "gradio", "torch", "transformers"])
 import re
 import gradio as gr
 import torch
 import transformers
+import requests
 
 import json
 from transformers import GPT2LMHeadModel, GPT2Tokenizer
@@ -13,12 +14,18 @@ device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 # Define a function for generating text based on a prompt using the fine-tuned GPT-2 model and the tokenizer
 def generate_text(prompt, length=100, theme=None, **kwargs):
 
-    model_url = "https://huggingface.co/spaces/sailormars18/Yelp-reviews-usingGPT2/blob/main/pytorch_model.bin"
-    config_url = "https://huggingface.co/spaces/sailormars18/Yelp-reviews-usingGPT2/blob/main/config.json"
-    generation_config_url = "https://huggingface.co/spaces/sailormars18/Yelp-reviews-usingGPT2/blob/main/generation_config.json"
+    model_url = "https://huggingface.co/spaces/sailormars18/Yelp-reviews-usingGPT2/resolve/main/pytorch_model.bin"
+    config_url = "https://huggingface.co/spaces/sailormars18/Yelp-reviews-usingGPT2/resolve/main/config.json"
     
-    # Load the model from the Hugging Face space
-    model = transformers.GPT2LMHeadModel.from_pretrained(model_url, config=config_url).to(device)
+    # Download the model and configuration files
+    model_path = "./pytorch_model.bin"
+    config_path = "./config.json"
+    
+    download_file(model_url, model_path)
+    download_file(config_url, config_path)
+    
+    # Load the model from the downloaded files
+    model = transformers.GPT2LMHeadModel.from_pretrained(model_path, config=config_path).to(device)
     
     # Load the tokenizer from the Hugging Face space
     tokenizer = transformers.GPT2Tokenizer.from_pretrained('gpt2')
@@ -63,6 +70,13 @@ def generate_text(prompt, length=100, theme=None, **kwargs):
 
     return generated_text
 
+def download_file(url, save_path):
+    response = requests.get(url, stream=True)
+    response.raise_for_status()
+    with open(save_path, "wb") as f:
+        for chunk in response.iter_content(chunk_size=8192):
+            f.write(chunk)
+
 # Define a Gradio interface for the generate_text function
 iface = gr.Interface(
     fn=generate_text,